高频雷达海杂波的多重分形建模方法研究

Multifractal modeling method for HF radar sea clutter

将多重分形理论引入高频雷达海杂波建模,提出了一种新的高频雷达海杂波的时域多重分形模型。对比分析了实测数据海杂波、分数布朗运动(FBM)模型和多重分形模型的质量指数函数和奇异指数,结果表明:该多重分形模型与实测数据具有相似的多重分形性。此外,通过对韦布尔分布、对数正态分布和瑞利分布三种最常用海杂波的概率密度函数的比较分析,并利用修正的柯尔莫哥洛夫-斯米尔洛夫(K-S)统计检验,得出了高频雷达海杂波的多重分形模型具有先验的统计特性的结论,从而进一步证明了该多重分形模型的合理性和有效性。

Multifractal theory is introduced for modeling HF radar sea clutter, and a new time-domain model based on multifractal for HF sea clutter is presented. Simulation analysis indicates that the multifractal model has a similar multifractality with real sea clutter data by comparison among multifractal model, real sea clutter data and Fractional Brownian Motion（FBM） model. In addition, by using probability density function comparison analysis among 3 most commonly used distributions such as Weibull, Log-Normal and Rayleigh distribution, and through a statistical test by modified Kolmogorov- Smirnov（K-S）, the conclusion of the proposed multifractal model of HF radar sea clutter possesses the apriori statistical characteristics is obtained, which further proves that this time-domain multifractal model for HF radar sea clutter is quite reasonable and effective.